COVID-19 Latin American fall and rebound since the - CEMLA
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Latin American fall and rebound since the COVID-19 Luciano Campos1 Danilo Leiva-León2 Steven Zapata2 1 Universidad de Alcalá 2 Banco de España 3 Ministerio de Hacienda de Colombia July 2021 The views expressed in this presentation do not necessarily reflect the views of the Banco de España, the Eurosystem or the Ministerio de Hacienda de Colombia.
Introduction I The COVID-19 crisis has significantly challenged the measurement of economic conditions around the globe I Latin America is not the exception BRASIL MEXICO PERU 6 Quarterly GDP Growth Before COVID-19 4 2 0 -2 -4 -6 -8 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 30 Quarterly GDP Growth Since COVID-19 20 10 0 -10 -20 -30 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020
Introduction I Context: During COVID-19 times a continuous reading of the Latin American economy’s vital sign is of the utmost importance I Aim 1:Tracking turning points in the region under this unstable environment I Aim 2: Tracking how deep (buoyant) an unfolding recessions (expansion) in the region can become I Problem: The magnitudes of real activity have become more heterogeneous than ever I This significantly complicates economic modelling
Literature I Hamilton (1989): Univariate Markov-switching models where unconditional means µexp and µrec are constant over time I Chauvet (1998): Multivariate Markov-switching (dynamic factor) models where unconditional means also constant
Literature I Hamilton (1989): Univariate Markov-switching models where unconditional means µexp and µrec are constant over time I Chauvet (1998): Multivariate Markov-switching (dynamic factor) models where unconditional means also constant
Literature I Hamilton (1989): Univariate Markov-switching models where unconditional means µexp and µrec are constant over time I Chauvet (1998): Multivariate Markov-switching models where unconditional means also constant I Eo and Kim (2016): Univariate Markov-switching models where unconditional means are time-varying I Leiva-Leon et al. (2021): Multivariate Markov-switching models where unconditional means are time-varying
What we do I We employ the framework recently proposed in Leiva-Leon et al. (2021) for the study of LATAM’s cyclical position: 1. Markov-Switching Dynamic Factor Model: Accounts for business cycle asymmetries 2. Mixed Frequency and Ragged Edges: Provides high-frequency real-time updates 3. Flexible Time-varying Means: Measures heterogeneous recessions and expansions I Estimate the proposed nonlinear factor model for the largest LATAM economies, using quarterly GDP and monthly activity indicators I Argentina, Brazil, Chile, Colombia, Ecuador, Mexico and Peru. I These inferences are summarized into a Latin American Weakness Index (LAWI) that provides: → timely assessments of LATAM’s cyclical position
What we do I We employ the framework recently proposed in Leiva-Leon et al. (2021) for the study of LATAM’s cyclical position: 1. Markov-Switching Dynamic Factor Model: Accounts for business cycle asymmetries 2. Mixed Frequency and Ragged Edges: Provides high-frequency real-time updates 3. Flexible Time-varying Means: Measures heterogeneous recessions and expansions I Estimate the proposed nonlinear factor model for the largest LATAM economies, using quarterly GDP and monthly activity indicators I Argentina, Brazil, Chile, Colombia, Ecuador, Mexico and Peru. I These inferences are summarized into a Latin American Weakness Index (LAWI) that provides: → timely assessments of LATAM’s cyclical position
What we do I We employ the framework recently proposed in Leiva-Leon et al. (2021) for the study of LATAM’s cyclical position: 1. Markov-Switching Dynamic Factor Model: Accounts for business cycle asymmetries 2. Mixed Frequency and Ragged Edges: Provides high-frequency real-time updates 3. Flexible Time-varying Means: Measures heterogeneous recessions and expansions I Estimate the proposed nonlinear factor model for the largest LATAM economies, using quarterly GDP and monthly activity indicators I Argentina, Brazil, Chile, Colombia, Ecuador, Mexico and Peru. I These inferences are summarized into a Latin American Weakness Index (LAWI) that provides: → timely assessments of LATAM’s cyclical position
The Model I Real activity indicators, yi,t , can be expressed as: yi,t = γi Θ(ft ) + ui,t , I The idiosyncratic components, ui,t , are given by ui,t = ψi,1 ui,t−1 + ... + ψi,P ui,t−p + ei,t , ei,t ∼ N (0, σi2 ) i.i.d. I The common factor, ft , follows flexible nonlinear dynamics to accommodate for recessions and expansions of different magnitudes, ft = µ0,τi (1 − st ) + µ1,τi st + ef ,t , ef ,t ∼ N (0, σf2 ) i.i.d., where st ∈ {0, 1}, with 0 for recessions, and 1 for expansions. I The regime-dependent means, µ1,τi and µ0,τi , correspond to the expected value of the factor, ft , during the τi−th expansion or the τi−th recession.
Data I We fit the nonlinear factor model to the seven largest economies in the LATAM region by employing the following quarterly and monthly indicators of activity:
Argentina Quarterly Real GDP Growth 10 0 -10 -20 1995 2000 2005 2010 2015 2020 Monthly Activity Index 10 0 -10 1995 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 1995 2000 2005 2010 2015 2020
Brazil Quarterly Real GDP Growth 10 0 -10 2000 2005 2010 2015 2020 Monthly Activity Index 0 -5 -10 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 2000 2005 2010 2015 2020
Chile Quarterly Real GDP Growth 10 0 -10 2000 2005 2010 2015 2020 Monthly Activity Index 5 0 -5 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 2000 2005 2010 2015 2020
Colombia Quarterly Real GDP Growth 10 0 -10 -20 1995 2000 2005 2010 2015 2020 Monthly Activity Index 5 0 -5 -10 1995 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 1995 2000 2005 2010 2015 2020
Ecuador Quarterly Real GDP Growth 0 -5 -10 1995 2000 2005 2010 2015 2020 Monthly Activity Index 5 0 -5 1995 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 1995 2000 2005 2010 2015 2020
Mexico Quarterly Real GDP Growth 10 0 -10 -20 1995 2000 2005 2010 2015 2020 Monthly Activity Index 10 0 -10 -20 1995 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 1995 2000 2005 2010 2015 2020
Peru Quarterly Real GDP Growth 20 0 -20 2000 2005 2010 2015 2020 Monthly Activity Index 10 0 -10 2000 2005 2010 2015 2020 Monthly Recession Probability 1 0.5 0 2000 2005 2010 2015 2020
Latin American Weakness Index I LAWI: proportion of the LATAM economy in recession K (l) X (l) LAWIt = ωκ,t sκ,t , κ=1 1 0.5 0 2000 2005 2010 2015 2020 Argentina Brazil Chile Colombia Ecuador Mexico Peru 1 0.5 0 2000 2005 2010 2015 2020
Latin American Weakness Index: Zoom into COVID crisis I The country-specific contributions to the LATAM economic weakness have exhibited significant variations in recent times I Monitoring the internal sources of economic weakness in the region represents is crucial for policy makers 1 0.5 0 2020.2 2020.4 2020.6 2020.8 2021 2021.2 2021.4 Argentina Brazil Chile Colombia Ecuador Mexico Peru 1 0.5 0 2020.2 2020.4 2020.6 2020.8 2021 2021.2 2021.4
Country-specific “Intensity” of Growth t f = (µ0,τi (1 − st ) + µ1,τi st ) + ef ,t , ef ,t ∼ N (0, σf2 ) |{z} | {z } |{z} Factor Intensity : µt Disturbances Argentina Brazil Chile Colombia 5 4 3 2 2 2 1 0 0 0 -2 0 -2 -1 -4 -4 -5 -2 -6 -3 -6 -8 -4 -8 -10 -10 -5 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 Ecuador Mexico Peru 6 5 4 4 2 2 0 0 0 -2 -5 -2 -4 -6 -4 -10 -8 -6 -10 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020
Latin American Intensity Index I For each country, the common factor can be decomposed into t f = (µ0,τi (1 − st ) + µ1,τi st )+ ef ,t , ef ,t ∼ N (0, σf2 ) |{z} | {z } |{z} Factor Intensity : µt Disturbances I LAII: Growth intensity of the LATAM economy K (l) X (l) LAIIt = ωκ,t µκ,t , κ=1 1 0.5 0 -0.5 -1 -1.5 -2 2000 2005 2010 2015 2020
Application: Effect of US Financial Conditions on LATAM I What is the time-varying relationship between U.S. financial conditions and LATAM’s economic weakness? LAIIt = αt + βt FCIt + et , I where αt and βt are assumed to follow random walks 2.5 Latin American Intensity Index U.S. Financial Conditions Index 2 1.5 1 0.5 0 -0.5 -1 -1.5 2000 2005 2010 2015 2020
Application: Effect of US Financial Conditions on LATAM I The contemporaneous correlation between LATAM real activity and FCI have increased substantially since the COVID crisis. I This estimates suggest the increasing importance of U.S. financial conditions for LATAM’s economic recovery 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 2000 2005 2010 2015 2020
Country-specific “Intensity” of Growth I There is heterogeneity regarding the sensitivity of country-specific weakness to U.S. financial conditions. 0.5 0 -0.5 -1 -1.5 -2 -2.5 -3 -3.5 -4 2018 2019 2020 2021 Argentina Brazil Chile Colombia Ecuador Mexico Peru
Conclusions I This paper brings new statistical measures for LATAM economies to the table for the use of policy makers 1. Latin American Weakness Index (LAWI): measures the share of LATAM region in recession 2. Latin American Intensity Index (LAII): measures the intensity of LATAM crisis and recoveries I The framework provides high frequency real-time updates on the cyclical position of LATAM economies as new information is released I We also asses the effect of external developments on the LATAM region, by focusing on the time-varying effect of U.S. financial conditions. I The estimates suggest that sensitivity of LATAM activity to U.S. financial conditions has substantially increased since the COVID-19 crisis.
You can also read